I'm Sam - engineer and fine-tuning practitioner. Previously I was founding engineer at Brev.dev working on building a GPU cloud. I studied computer science and electrical engineering at Imperial College London.

This blog is a means for me to learn deeply about topics that interest me, namely RAG and fine-tuning. I believe the RAG vs fine-tuning debate is pointless. You really need both to make a model truly useful: RAG gives a model new knowledge and fine-tuning gives a model new skills.

My main project at the moment is a private and open-source AI writing and knowledge management desktop app: available here.